Prospective analysis of alternative serv”>American Journal of Emergency Medicine 38 (2020) 624-628

Contents lists available at ScienceDirect

American Journal of Emergency Medicine

journal homepage:

Prospective analysis of alternative services and cost savings of avoidable admissions from the ED

Taketo Watase ?, Karl Jablonowski, Amber Sabbatini

Department of Emergency Medicine, University of Washington, 325 9th Ave, Seattle, WA 98104, USA

Introduction

Each year, nearly $1.1 trillion - or a third of national health care expenditures - is spent on hospital care [1]. Wide variation in admission practices across individual providers and hospitals exists [2] and is often cited by payers and policymakers as evidence that many hospital admissions are unnecessary. More than 80% of unscheduled hospital admission decisions are made in the emer- gency department (ED) [3], yet, to date, there is paucity of research examining potentially avoidable admissions from the ED.

The Centers for Medicare & Medicaid Services (CMS) defines potentially avoidable hospitalizations as hospitalizations that either (1) may have been prevented with better ambulatory care or (2) treated effectively outside of an inpatient setting [4]. Prior work has generally focused on estimating the rate of potentially avoidable hospitalizations from retrospective data through the use of discharge diagnoses. For example, a common strategy used by both researchers and policymakers has been to examine rates of hospitalizations for ambulatory care sensitive conditions (ACSC) [5-8]. ACSC hospitalizations capture the first definition of a poten- tially avoidable hospitalization; they are a group of diagnoses for chronic conditions that are thought to be preventable with the pro- vision of timely and effective ambulatory care. By the time these patients arrive to the ED, they are suffering complications of their disease that requires Urgent treatment, including hospitalization. While the opportunity to prevent hospitalization through better ambulatory care has passed, hospitalization may still be avoided for many of these patients if safe and effective alternative care pathways are readily available. Therefore, this is an area high- lighted by CMS’s second definition of avoidable admission where enhanced acute care and wrap-around services for ED patients could be developed to further optimize hospitalization decision in the ED.

The goal of this study was to determine, in real time, the

proportion of ED admissions that emergency physicians identify as potentially avoidable and the alternative services that would have been necessary to prevent those admissions. We further

* Corresponding author.

E-mail address: [email protected] (T. Watase).

estimated the net annual cost-savings that would be associated with avoidance of these admissions.

Methods

This was a prospective study conducted over a ten-week period at an academic, urban, county hospital with 64,000 visits per year. Unscheduled adult medical and surgical hospitalizations from the ED were eligible for study. We excluded patients admitted to the ICU, transfers from another hospital, and major Trauma activations, as these hospitalizations are unlikely to be at the discretion of the ED provider. For other admissions, a questionnaire (Appendix A1) was handed to the attending ED physician by the charge nurse at the time a decision to admit was made. Physicians were asked whether the admission could have been avoided had some alterna- tive service been readily available to them at the time of disposi- tion. For admissions felt to be avoidable, physicians were provided with a checklist to specify the specific service that would have been required to safely discharge the patient, including treat- ment in an ED-basED observation unit (EDOU), expedited clinic follow-up with a primary care provider or specialist (asked to spec- ify timeframe), direct transfer to a skilled nursing facility (SNF), home health services, an expedited outpatient diagnostic test (e.g. stress test, MRI; asked to specify), or other (asked to specify). Finally, clinicians were also asked to list any social factors that may have contributed to their decision to hospitalize.

Characteristics of admissions identified as potentially avoidable were compared to those not considered avoidable. We define a potentially avoidable admission as one in which an attending emergency physician would have discharged the patient if some alternative service were readily available and there were no con- comitant social factors precluding safe discharge. For each admis- sion identified as potentially avoidable by emergency providers, we abstracted demographic (age, sex, comorbidities, primary payer) and clinical information (chief complaint, triage level, diag- nosis from ED, hospital length of stay , hospital discharge diagnosis, escalation of care to ICU, discharge disposition, and whether the admission was ultimately billed as an inpatient admission versus observation stay). We calculated the proportion of potentially avoidable hospitalizations attributed to each class of alternative service from our survey responses, and applied these

https://doi.org/10.1016/j.ajem.2019.11.001

0735-6757/(C) 2019

to annual ED visit data to estimate the number of hospitalizations that could be potentially avoided each year.

Table 2

Health Services that Potentially Avoided Admissions and its Cost Savings.

Resource inputs for our cost-savings model are shown in

Appendix Table A2 & A3. Cost-savings were estimated from the

Services Avoided Admissions (%)

Cost Savings (k

$)

hospital perspective with service costs obtained from our health systems’ hospital billing department. These included average daily hospital cost for an observation stay and inpatient admission, aver- age costs for an outpatient primary care visit, and average costs for

ED^& observation unit 54 (43.9%) 1,919

Direct transfer to SNF* 25 (20.3%) 2,678

Expedited specialty clinic f/u^% 14 (11.4%) 329

Expedited primary clinic f/u 7 (5.7%) 224

Home health services 6 (4.9%) 838

each specialty clinic and outpatient diagnostic test identified as necessary by physicians. hospital costs associated with each potentially avoidable hospitalization were estimated by multiply- ing the hospital LOS in days for each avoidable admission with the average daily cost for either an observation stay or inpatient admission, depending on how the hospitalization was billed. Then, we calculated any costs that would be accrued in lieu of hospital- ization from the alternative services identified in our physician surveys. The cost savings for each potentially avoidable hospital- ization represented the difference between hospital costs saved and these accrued ambulatory costs. Annualized cost-savings were

Expedited outpatient diagnostics

^& ED: Emergency Department.

Other	9 (7.3%)	247
Total	123 (100%)	6,306

^* SNF: Skilled Nursing Facility.

^% f/u: follow-up.

4. Discussion

8 (6.5%) 67

calculated by multiplying the estimated number of potentially avoidable admissions by the average cost difference across each alternative service category.

3. Results

Physician surveys were collected for 918 hospitalizations meet- ing our inclusion criteria. Of these, 123 were considered to be potentially avoidable (13.4%). Patients with potentially avoidable hospitalizations were more likely to be older, female, and higher rates of substance use, psychiatric illness, and Public insurance coverage (Table 1). Access to an EDOU (43.9%), a pathway for direct transfer to a SNF (20.3%) and specialty clinic follow-up within 24-72 h (11.4%) were the top 3 alternative services identified by physicians, and comprised nearly three-quarters of avoidable hos- pitalizations. Expedited outpatient diagnostic testing, expedited primary care follow-up, and home health services attributed the remainder of the responses (Table 2).

At our institution, we estimate that a total of 1034 hospitaliza- tions could potentially be avoided each year, amounting to a cost- savings of approximately $6.3 million annually, should the hospital invest in alternative service pathways and make them readily available to ED providers. The three alternative services with the highest potential savings were direct transfer to SNF, EDOU, and home health services, saving $2.68 million, $1.92 million, and

$0.84 million, respectively, contributing to 86% of potential annual savings (Table 2).

Table 1

Characteristics of surveyed patients.

In this study, we surveyed emergency physicians in real time to estimate the rate of potentially avoidable hospitalizations at an academic county hospital, determine the alternative services that would have been necessary to avoid these admissions, and poten- tial cost savings to the health system. We found that nearly 1 in 7 non-critical care hospitalizations could have been avoided if alter- native services were readily available to ED providers at the time of disposition. Physicians most commonly cited access to an EDOU, availability of a pathway permitting direct transfer to a SNF, and expedited subspecialty follow-up (with neurology, cardiology and orthopedics) as key services, which could have avoided over 75% of these admissions. We estimate the development of enhanced clinical pathways for acute and Emergent care in our system alone would result a cost-savings of $6.3 million per year.

Our study adds to the body of literature on avoidable admis- sions in three ways. First, we estimate the rate of potentially avoid- able hospitalizations for an ED population. Despite the ED being the primary source of unscheduled hospitalizations, few studies have focused on the proportion of emergency admissions that are avoidable. Second, we demonstrate the feasibility of using a prospective method for capturing these potentially avoidable admissions. This compares to the far more common approach of using retrospective data and relying on blunt Diagnostic categories like ACSCs, which may not reflect the clinical needs and hetero- geneity of disease observed for specific patients within diagnostic groups. Other studies that have not classified potentially avoidable admissions using Diagnosis codes, have nevertheless relied on ret- rospective approaches, such as asking hospitalists and primary care physicians their opinion of whether an admission could have been avoided after retrospective chart review [9]. To our knowl-

edge, this study is the first to prospectively estimate the rate of

Characteristic Avoidable N(%) (N = 123)

Age

Not Avoidable N(%) (N = 795)

avoidable hospitalizations from the ED by querying emergency physicians in real time, an approach that could be utilized in other

health systems for quality improvement and to assess service

<18 5 (4.1) 43 (5.4)

18-39 24 (19.5) 195 (24.5)

40-64 59 (48.0) 376 (47.3)

>65 35 (28.5) 181 (22.8)

Female 57 (46.3) 251 (31.6)

>= 2 Comorbidities 58 (47.2) 368 (46.3)

Psychiatric Illness 16 (13.0) 74 (9.3)

needs. Third, we employ CMS’ second definition of a potentially avoidable hospitalization, or hospitalizations could have feasibly been treated in an ambulatory setting had appropriate services been available, in order to highlight the potential of ED-focused interventions to reduce hospital costs.

Substance Abuse	28 (22.8)	158 (19.9)	The way that potentially avoidable hospitalizations have been
non-English language Payer	15 (12.2)	135 (17.0)	characterized in both research and policy has implications for the strategies that are implemented to reduce hospital care. To date,

these strategies have largely focused on strengthening primary care and coordinating chronic disease management with the pri- mary goal of preventing ED visits and hospital care [10,11]. Simi- larly, there has been intense focus, largely driven by recent payer policies, on improving the quality of care transitions and

Commercial	10 (8.1)	150 (18.9)
Medicaid	53 (43.1)	277 (34.8)
Medicare	49 (39.8)	256 (32.2)
Uninsured	7 (5.7)	77 (9.7)
Other	0	5 (0.6)
Unknown	4 (3.3)	30 (3.8)

post-acute care among inpatients being discharged from the hospi- tal [12], which targets secondary prevention of recurrent ED visits or readmissions. However, to date, little attention has been given to ED-based strategies that can reduce avoidable hospitalizations, despite the fact that ED visits continue to increase across all payer groups each year [13].

An ED provider’s decision to hospitalize a patient weighs a vari- ety of objective clinical factors, subjective clinical factors, provider or patient preferences [14,15], economic, psychosocial concerns [16,17,18,19], and operational concerns [20]. Inherent in balancing all of these factors and allowing for a safe discharge is the availabil- ity of timely and effective alternatives to hospitalization that are readily available to providers at the time of ED evaluation. Yet, because health systems have not often developed such alternative systems of care, ED providers frequently hospitalize patients when uncertainty exists. The true rate of potentially avoidable hospital- izations reflects some combination of those preventable by inter- ventions that strengthen primary care (keeping patients healthy) and those avoided by developing enhanced acute care services at the point of ED care. In order to save costs, especially in an era of increasing health system consolidation and integration, compli- mentary strategies targeting both primary care and acute alterna- tives to admission at the point of ED care are necessary.

With greater movement towards Value-based care, incentives around hospitalization are rapidly changing. Recent policies have discouraged short inpatient admissions in favor of observation stays [21] and penalized hospitals for readmissions [22,23]. Fur- ther, hospitals and physician groups are entering into Alternative Payment Models with both public and private payers wherein they assume greater financial risk in providing a continuum of care for a population of patients. These changing payment policies provide strong incentives to health systems to find innovative ways to keep patients healthy and out of the hospital. Our work highlights several key acute care service pathways, which can readily be implemented in many health systems to reduce unnecessary or low-value hospitalizations.

Our results must be interpreted in light of the following limita- tions. Due to its single-center nature, the results of this study may not be generalizable to other institutions. The included patient cases were convenience samples of all ED admissions, therefore not all admissions were captured in this study. However, we note that demographic and clinical characteristics of the hospitaliza- tions captured in the study were similar to those that we did not have surveys for (Table A4). Finally, the determination of whether an admission was avoidable or not was ultimately subjective and based on the provider’s opinion at the time of admission. It may be difficult for providers to imagine whether a patient could safely be cared for in an alternative venue when they have no experience with those alternative pathways, and thus our results may actually be underestimates of the number of potentially avoidable admissions.

Conclusion

Nearly 1 in 7 unscheduled admissions from the ED may be avoidable if hospitals develop enhanced service pathways for pop- ulation health available to ED providers. Emergency providers identified EDOUs, direct transfer to SNF and timely specialty follow-up (e.g. within 24-72 h) as the alternatives to admission that would yield the highest reductions in ED hospitalizations.

Source of support

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Presentation

Presented at Society of Academic Emergency Medicine Annual Meeting, May 19, 2017, Orlando, FL

Appendix

A1. Questionnaire handed to attending physicians at time of admission

See Tables A2-A4.

Table A2 Utilization inputs for cost analysis.
	Proportion of Admissions Identified as Avoidable^	Estimated Annual Admissions	Median Hospital LOS (days)	Observation Stay (%)	Inpatient Admission (%)
ED Observation Unit	0.0331	454	2.1	0.26	0.74
Skilled Nursing Facility	0.0153	210	4.4	0.29	0.71
Diagnostic Test	0.0086	118	0.69	1.00	0.00
Expedited Specialty Care	0.0049	67	1.07	0.17	0.83
Home Health Care	0.0043	59	4.9	0.29	0.71
Expedited Primary Care	0.0037	50	1.18	0.17	0.83
Other	0.0055	76	1.16	0.35	0.65
Total Avoidable	0.0754	1034
^ From survey.

Table A3

Cost inputs from hospital billing data.

Costs ($)

Ambulatory care-sensitive conditions. Annals Family Med 2013;11(4):363-70. https://doi.org/10.1370/afm.1498.

[7] Centers for Medicare & Medicaid Services; ”2016 Measure Information About the Hospital Admissions for Acute and Chronic Ambulatory Care-Sensitive Condition (ACSC) Composite Measures, Calculated for the 2018 Value-Based

Daily Observation	$1,953	Payment Modifier Program”, Accessed August 6, 2019 from <https://www.
Daily Inpatient (non ICU)	$3,289	cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeedback
Primary care clinic	$288	Program/Downloads/2016-ACSC-MIF.pdf>.
Cardiology Clinic	$346	[8] Hodgson K, Deeny SR, Steventon A. ambulatory care-sensitive conditions: their
Orthopedics Clinic	$340	potential uses and limitations. BMJ Qual Saf. 2019 Jun;28(6):429-33.
Neurosurgery Clinic	$714	[9] Soulen JL, Duggan AK, DeAngelis CD. Identification of potentially avoidable
general surgery clinic	$859	pediatric hospital use: admitting physician judgment as a complement to

Table A4

Gastroenterology Clinic $320

Ophthalmology Clinic $341

Avg. Specialty Clinic $349

Outpatient Stress Echo $478 Outpatient Venous Duplex $393 Outpatient Endoscopy $1,143

utilization review. Pediatrics.1994 Oct;94(4 Pt 1):421-4.

Clarke JL, Bourn S, Skoufalos A, Beck EH, Castillo DJ. An innovative approach to Health care delivery for patients with chronic conditions. Popul Health Manag 2017 Feb;20(1):23-30.

Centers for Medicare & Medicaid Services, ”Chronic Care Management Services”, Accessed August 8, 2019 from <https://www.cms.gov/Outreach- and-Education/Medicare-Learning-Network-MLN/MLNProducts/Downloads/ ChronicCareManagement.pdf>.

Rennke S, Nguyen OK, Shoeb MH, Magan Y, Wachter RM, Ranji SR. Hospital- initiated transitional care interventions as a patient safety strategy: a systematic review. Ann Intern Med. 2013 Mar 5;158(5 Pt 2):433-40.

Patient demographics of Surveyed vs Not-surveyed Groups.

	Surveyed	Not-surveyed
Number	1629*	1988
Age	49.1	49.5
Male (%)	66.3	66.9
Median ED Length of Stay (hrs)	5.7	5.7
Median Hospital Length of Stay (days) Payer (%)	3.2	2.9
Medicaid	549 (33.7)	721 (36.3)
Medicare	513 (31.5)	561 (28.2)
Commercial	342 (21)	427 (21.5)
Self-Pay	146 (9)	179 (9)
Other	18 (1.1)	16 (0.8)
Blank	61 (3.7)	84 (4.2)

^* Total number of patients including those meeting exclusion criteria.

References

1. Centers for Medicare & Medicaid Services. ”National health expenditures 2017 Highlights.” Accessed July 30, 2019 from <https://www.cms.gov/Research- Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealth ExpendData/downloads/highlights.pdf>.
2. Sabbatini AK, Nallamothu BK, Kocher KE. Reducing variation in hospital admissions from the emergency department for low-mortality conditions may produce savings. Health Aff (Millwood). 2014 Sep;33(9):1655-63.
3. Kocher KE, Dimick JB, Nallamothu BK. Changes in the source of unscheduled hospitalizations in the United States. Med Care. 2013 Aug;51 (8):689-98.
4. Segal M, Rollins E, Hodges K, Roozeboom M. Medicare-Medicaid eligible beneficiaries and potentially avoidable hospitalizations. Medicare Medicaid Res Rev. 2014 Jan 15;4(1).
5. Solberg LI, Ohnsorg KA, Parker ED, Ferguson R, Magnan S, Whitebird RR, et al. Potentially preventable hospital and emergency department events: lessons from a large innovation project. Perm J 2018;22:17-102.
6. Freund T, Campbell SM, Geissler S, Kunz CU, Mahler C, Peters-Klimm F, Szecsenyi J. Strategies for reducing potentially avoidable hospitalizations for
7. Lin MP, Baker O, Richardson LD, Schuur JD. Trends in emergency department visits and admission rates among US Acute Care Hospitals. JAMA Intern Med. 2018 Dec 1;178(12):1708-10.
8. Halm EA, Atlas SJ, Borowsky LH, Benzer TI, Metlay JP, Chang YC, et al. Understanding physician adherence with a pneumonia practice guideline: effects of patient, system, and physician factors. Arch Intern Med 2000 Jan 10;160(1):98-104.
9. Aujesky D, McCausland JB, Whittle J, Obrosky DS, Yealy DM, Fine MJ. Reasons why emergency department providers do not rely on the pneumonia severity index to determine the initial site of treatment for patients with pneumonia. Clin Infect Dis. 2009;15;49(10):e100-8.
10. Kuder JM, Vilmain JA, Demlo LK. Exploring physician responses to patients’ extramedical characteristics. The decision to hospitalize. Med Care 1987;25 (9):882-93.
11. Rosenthal GE, Harper DL, Shah A, Covinsky KE. A regional evaluation of variation in low-severity hospital admissions. JGIM 1997;12(7):416-22.
12. Goss CH, Rubenfeld GD, Park DR, Sherbin VL, Goodman MS, Root RK. Cost and incidence of social comorbidities in low-risk patients with community- acquired pneumonia admitted to a public hospital. Chest 2003;124 (6):2148-55.
13. Labarere J, Stone RA, Obrosky DS. Factors associated with the hospitalization of low-risk patients with community-acquired pneumonia in a cluster- randomized trial. JGIM 2006;21(7):745-52.
14. Calder LA, Forster AJ, Stiell IG, Carr LK, Perry JJ, Vaillancourt C, et al. Mapping out the emergency department Disposition decision for High-acuity patients. Ann Emerg Med 2012;60(5):567-576.e4.
15. Lind KD, Noel-Miller CM, Sangaralingham LR, Shah ND, Hess EP, Morin P, Fernanda Bellolio M. Increasing Trends in the Use of hospital observation Services for Older Medicare Advantage and Privately Insured Patients. Med Care Res Rev. 2019 Apr;76 (2):229-39.
16. MedPAC June 2015 Report to the Congress: Medicare and the Health Care Delivery System, Chap 7, Accessed September 4, 2019 from <http://www. medpac.gov/docs/default-source/reports/chapter-7-hospital-short-stay-policy- issues-june-2015-report-.pdf?sfvrsn=0>.
17. Centers for Medicare and Medicaid Services. Baltimore, MD, Accessed September 4, 2019 from <http://www.cms.gov/Medicare/Medicare-Fee-for- Service-Payment/AcuteInpatientPPS/Readmissions-Reduction-Program.html>.